How to quieten electrical noise on DC motor?

Discussion in 'The Projects Forum' started by tracecom, Feb 29, 2016.

  1. tracecom

    Thread Starter AAC Fanatic!

    Apr 16, 2010
    3,869
    1,393
    I have a microcontroller that is feeding an H-bridge in order to control the on/off and direction of a small 6VDC motor. I have two LEDs connected back to back across the H-bridge output in order to have a visual indication of the output polarity from the H-bridge. Everything works well until I connect the motor, at which time the motor buzzes but doesn't turn and both the LEDs light as if the polarity is rapidly switching. I suspect the uC is being interfered with by electrical noise from the motor. I have a .1 uF cap across power legs of the uC, but I need to do more.

    Suggestions are welcome.

    Thanks.
     
  2. ian field

    Distinguished Member

    Oct 27, 2012
    4,413
    782
    Could be anything - and without more info, we'll never know.
     
  3. shortbus

    AAC Fanatic!

    Sep 30, 2009
    4,000
    1,512
    Isn't this why they make "motor shields"? A type of breakout board to isolate the motor from the uC.
     
  4. tracecom

    Thread Starter AAC Fanatic!

    Apr 16, 2010
    3,869
    1,393
    I added .1uF caps from each motor terminal to the case and another from one motor terminal to the other. Still no joy.
     
  5. Alec_t

    AAC Fanatic!

    Sep 17, 2013
    5,766
    1,099
    Do you have a star ground system so that motor current doesn't affect signal current paths?
    Are the power leads to the motor a shielded twisted pair?
     
  6. tracecom

    Thread Starter AAC Fanatic!

    Apr 16, 2010
    3,869
    1,393
    Yes

    and No

    I will unsolder the leads on the control circuit and move them so that they run straight away from the uC.
     
  7. Kermit2

    AAC Fanatic!

    Feb 5, 2010
    3,766
    928
    Does the circuit use the same power source as the motor?
     
  8. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,281
    1,231
    We could use some more information:
    Schematic, power supply, etc.
    Some possibilities:
    If there is no dead time the switches on the same side of the H can be on for a short time causing the supply to drop. Some dead time (leisure time) between direction changes and a big cap, say 220 Ufd. from the top of the H to the ground for the H.
    Flyback diodes?
    0.1Ufd. may be a little large for the brush noise caps. - maybe .01.
    Return the ground and the supply for the H directly to where the supply voltage enters the board and place a large cap there as well. Then power the micro from the same point.
    Pay attention to the micro reset pin.
     
    tracecom likes this.
  9. tracecom

    Thread Starter AAC Fanatic!

    Apr 16, 2010
    3,869
    1,393
    I don't have the actual motor that the circuit will ultimately be used with, so describing the motor I used wouldn't help. Tomorrow morning, I am picking up a duplicate of the "real" motor and will repeat my tests with it. I will post a schematic then as well. Thanks for the suggestions.
     
  10. tracecom

    Thread Starter AAC Fanatic!

    Apr 16, 2010
    3,869
    1,393
    Yes...but. In its final application, the circuit and the motor will both run from a set of four alkaline C cells. The motor will run directly on the 6V, but the control circuitry will run on 3.3V derived from the 6V via a linear regulator. Rather than batteries, I have been using a 7.5V regulated wall wart.
     
  11. tracecom

    Thread Starter AAC Fanatic!

    Apr 16, 2010
    3,869
    1,393
    Here is the schematic drawing. I rerouted the leads from the H-bridge away from the uC, but no change. I also eliminated the wall wart and used a battery for the 6V supply, but no change.

    I have never used this H-bridge IC before; I will pick up the actual motor to be used in a few hours and try again.

    RK Motor Controller Schematic.JPG
     
  12. GopherT

    AAC Fanatic!

    Nov 23, 2012
    5,979
    3,704
    Try putting isolating the motor power by putting a diode between the power (battery pack/adapter) and the H-bridge. Add a good sized electrolytic capacitor (470 to 1000uF) from diode's cathode to ground. That way, back EMF from motor is eliminated for logic level circuitry.

    Could you post a schematic of the rest of your project - it may not be a noise issue.
     
  13. tracecom

    Thread Starter AAC Fanatic!

    Apr 16, 2010
    3,869
    1,393
    Thanks for the input. The schematic shows the whole circuit; its purpose is simply to control the motor based on the operation of the five switches.

    ETA: Tried the diode and cap you suggested; no change. I am off to get the "real" motor.
     
    Last edited: Mar 1, 2016
  14. GopherT

    AAC Fanatic!

    Nov 23, 2012
    5,979
    3,704
    @tracecom
    Now I see the schematic.

    The motor driver datasheet has a recommendation about in-rush current limiting
    upload_2016-3-1_9-44-22.png

    Also, at 3.3V supply to pic-axe, outputs are near 2.5V. On the low end of I/O for your motor driver chip. You could try running the PICAxe on 5V to see if that reduces noise.

    Finally, You could try supplying the motor-control Chip's Vdd with a separate logic-level voltage (see optional Diode isolation - circled in Red) and Vm pin (highlited in yellow) for the motor power input. Motor noise may be screwing up the Vref pin signaling.


    upload_2016-3-1_9-50-36.png
     
  15. tracecom

    Thread Starter AAC Fanatic!

    Apr 16, 2010
    3,869
    1,393
    OK, I am back with the "real" motor and it works properly with my original circuit. That's good, but it still leaves me in a quandary: knowing that there is some problem, I don't know whether to assume it's exclusively with the motor that doesn't work or not. The circuit is not for production or even field trials; it's just for a couple of demo units.

    With regard to your markings on the datasheet page above, I am not using a voltage divider to control the speed...just a 10k to Vdd. I have a 10uF cap between Vm and ground. The datasheet wasn't too clear on where to connect Vm, but I found that it worked if I connected it to Vdd.
     
    Last edited: Mar 1, 2016
  16. GopherT

    AAC Fanatic!

    Nov 23, 2012
    5,979
    3,704

    Based on other motor controller chips I've used, there is usually a Vdd for logic level (up to 18 V if using CD4000 series, for example) but should be at what ever will produce the right switching for the two inputs. The datasheet does not show that the logic level switching thresholds changes with change in Vdd. I am confused, too.
     
Loading...